Income and wealth

It is necessary to examine measures of income at this stage as the issue of housing investment hinges on the financial resources of the households. In our consideration of incomes in this comparative study, we have followed Summers and Heston (1988) in the use of Purchasing Power Parity (PPP), which adjusts local costs through multiplying the official exchange rate by a factor related to the cost of living in that country to a hard currency, in our case pounds sterling.

Experience has shown that income and expenditure questions elicit very different data. Previous work in Ghana has consistently found that expenditure is stated to be an average of 2.4 times stated income (Malpezzi et al., 1990; Tipple, 1984). Obviously this cannot be so as, except in the very short term, no one can spend anything which is not income in some form but wages, gifts, bribes, produce in kind, business profits, etc., can all form parts of an aggregate income considerably higher than the bottom-line wage income stated to a gullible interviewer.

While it might be more correct to refer to expenditure in the analysis, we refer to income instead. This is because our expenditure data are standing for a measure of long-term income and most housing studies talk in terms of income relative to house cost, affordability, etc. We thus feel it would be valid for a reader to compare our data directly with income data from their own context, as long as such data contain all forms of income for all household members.

Regular savings, where a household lives below its income and puts money away for the future, are probably rare among low-income households. It is more likely that so-called savings are only deferred periodic spending on furniture, clothing, school fees, festivals, etc. The study collected data on both income and expenditure so the decision on the most useful could be made country by country. In the event, expenditure was universally used. We also have a wealth indicator as a back-up measure.

Data on the percentage of income spent on food are also presented, and are likely to be inversely proportional to income. Thus, they make a reasonable measure of wealth and ability to spend on housing to augment expenditure and the wealth indicator (Tipple et al., 1997).